Re: [boost] [smart_ptr] optimize current enable_shared_from_this with 100% compatibility

From: Frank Mori Hess (frank.hess_at_[hidden])
Date: 2008-04-23 15:18:32

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wednesday 23 April 2008 14:33 pm, Daniel Frey wrote:
> On Wed, 2008-04-23 at 08:45 -0400, Frank Mori Hess wrote:
> > The comment is obsolete, now that the pointer value is calculated in the
> > shared_from_this() calls.  Yes, the deleter_wrapper only needs a
> > shared_count.
>
> The comment suggest that the lazy initialization is just a work-around
> for the more desirable initialization during the ctor. Is this no longer
> true?

If you initialize in the constructor, you have to always create a 
deleter_wrapper.  Currently, one is only created if you need it.  I'd guess 
you would find the current behavior preferrable.

> > And what I was getting at before, _internal_shared_count could be made a
> > local variable in init_internal_shared_once(), if you add a line where
> > you stuff it inside its own deleter_wrapper.  Anywhere else it is used,
> > it can be obtained from _internal_weak_count.
>
> I think we should pick the low-hanging fruits first. If it's OK for you
> I'll apply the attached patch.

It's fine.

> For the vtable, we could remove it by switching shared_from_this from a
> member function to a free function that takes T* as a parameter, but
> that's of course a breaking change...

Didn't you propose something before like implementing the current esft 
interface as a derived class from a vtable-free base class that uses the free 
function approach?  The last thing I saw from Peter about that was it needed 
to maintain compatibility.

- -- 
Frank
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFID4uI5vihyNWuA4URAmToAKCUAc6X+P6FCXhPMYiDEQFgjDGE3gCcDYaS
keDGK/mqzefRjIvzOqwwbNU=
=qpxu
-----END PGP SIGNATURE-----